Hydrogen cyanide

Toxicological Summary

Workers - Hazard via inhalation route

Systemic effects

Long term exposure

Hazard assessment conclusion:

DNEL (Derived No Effect Level)

Value:

0.78 mg/m³

Most sensitive endpoint:

repeated dose toxicity

DNEL related information

Overall assessment factor (AF):

10

Modified dose descriptor starting point:

NOAEC

Acute/short term exposure

Most sensitive endpoint:

acute toxicity

DNEL related information

Local effects

Acute/short term exposure

DNEL related information

Workers - Hazard via dermal route

Systemic effects

Long term exposure

Hazard assessment conclusion:

DNEL (Derived No Effect Level)

Value:

0.054 mg/kg bw/day

DNEL related information

Overall assessment factor (AF):

10

Modified dose descriptor starting point:

NOAEL

Acute/short term exposure

Most sensitive endpoint:

acute toxicity

DNEL related information

Additional information - workers

The repeat dose toxicity of HCN was reviewed by ECETOC in the JACC report on Cyanides (No. 53, 2007) and it was concluded that an acceptable chronic level of exposure to CN- could be established using a weight-of-evidence approach that uses all the available repeat dose experimental studies in animals and human studies on the effects of exposure to cyanides/thiocyanate.

ECETOC proposed that “Taken together, the data regarding chronic toxicity mediated by thiocyanate from the key studies in animals and from human experience provide a relatively consistent picture. The repeated dose animal studies of Hébert (1993) and Monsanto (1984) appear to offer the most robust basis for determining a threshold for chronic toxicity in humans. The NOAELs in these studies were broadly consistent, i.e. 25.6 mg CN-/kgbw/d in mice, and 12.5 or 10.4 mg CN-/kgbw/d in rats. From the studies of Leuschner et al (1991) and Monsanto (1985) in rats and Tewe and Maner (1980) in pigs it can be deduced that serum thiocyanate levels of 345 to 475 μmol SCN-/l (20 - 80 μg/ml) consistently did not have goitrogenic effects on the thyroid if the diet was not iodine deficient. Those serum concentrations are very similar to the concentration of 345 to 690 μmol/l (20 - 40 μg/ml) that was regarded as a human NOAEL for thiocyanate therapy. Cliff et al (1986) found that levels of 250 μmol SCN-/l serum (14.5 μg SCN-/ml) did not lead to an increased incidence of goitre in a cassava eating population that was not iodine deficient.

Similar findings were reported by Banerjee (1997) for electroplating workers in which serum concentrations of 316 μmol/l (18 μg SCN-/ml) did not lead to goitre. Both of the latter populations had slight changes in thyroid hormone levels that can be considered as an adaptive response.”

The ECETOC review recognised that due to background variations in co-exposure to cyanogenic sources (e.g., environmental (combustion), dietary (cyanogenic foods) and habitual (i.e. smoking)) and levels of dietary iodine, it is not possible to distinguish precisely the effects of exogenous CN on thyroid hormone levels. For this reason, ECETOC selected the threshold for clinical diagnosis between adaptive and goitrogenic changes as a POD i.e. a NOAEL rather than a NOEL.

ECETOC went on to propose that “A level of 15 μg SCN-/ml can be used as a starting point to derive a safe concentration of cyanide for humans, using the following equation.

[CN-] mg/m3 = ([SCN-] mg/l × 0.25 l/kgbw × 70 kgbw × 26) divided by

(10 m3 × 0.5 × 0.8 × 3.9 × 58)

Where

0.25 l/kgbw = distribution volume of SCN

70 kg = body weight

26 = molecular weight CN

10 m3 = respiratory volume during an 8hour day

0.5 = absorption of cyanide by the inhalation route: 50%

0.8 = conversion of cyanide to thiocyanate: ~ 80%

3.9 = elimination of thiocyanate, half-life = 2.7 d, residence time (t0.5/ln 2) = 3.9 d

58 = molecular weight SCN-

The derived safe inhalation concentration of cyanide for humans exposed for 8 hours at the workplace of 7.5 mg/m3 can be used as a departure point for the calculation of DNELworker.

Assessment Factors:

In considering the overall level of uncertainty, typically the following aspects should be considered:

· interspecies differences;

· intraspecies differences;

· differences in duration of exposure;

· issues related to dose-response; and,

· quality of available human database including: completeness, consistency, reliability, healthy worker effect, study size.

Interspecies differences: In terms of intraspecies differences, it should be recognized that the POD, using the weight-of-evidence approach, considers both human as well as animal data. In the case of the human data, no interspecies assessment factor is necessary as the studies are already in humans. As such, the animal studies, albeit according to OECD guidelines, in more than one species, by different routes and sub-chronic in duration, merely provide additional confidence in the overall assessment in humans.

Intraspecies differences: As recognized previously, sensitive sub-populations would include individuals with insufficient dietary iodine, insufficient thiosulphate supply (e.g., in the case of malnutrition) or impaired renal function. Whereas the human studies relied upon for development of the POD typically included several hundred to several thousand individuals from as young as 5 to 80 years old, they were likely representative of the general population. As such, it is acknowledged that an intraspecies assessment factor may be viewed as necessary. Typically, a factor no greater than 10 is used in this regard. It is recognized that there is residual uncertainty regarding developmental effects due to iodine insufficiency and its impact on thyroid hormone levels and this is addressed under completeness of the database (below).

Differences in duration of exposure: In terms of differences in duration of exposure, the clinical studies by Barker et al. (1941) included 45 and 246 patients that received thiocyanate therapy for 1 to 4, or 4 to 10 years. The study in a cassava-eating population by Cliff et al., (1986) included 276 individuals (subset of 27 for some measurements) between 5 and 60 years old that had lived on this staple diet all their life (control group aged 19 – 85). The other supporting studies of Barrère et al. (2000, 2002) included 2987 subjects aged between 36 and 60 years; Bourdoux et al. (1978) included 66 individuals aged 9 to 16 yrs, 386 aged 0 to 80 yrs and 988 aged 15 to 30 yrs; Jackson (1988) studied 73 individuals aged 15 to 80 yrs; Knudsen et al., (2000, 2002) involved 2656 individuals aged from 41 to 71years and Dahlberg et al., 1984 included 37 individual 16 to 54 yrs. From the multitude of studies, it should be readily apparent that the duration of exposure in the human population studies was sufficient to address any latency of the critical effect.

Issues related to dose-response: The dose response of the critical effect (goitre) is relatively well known in terms that it comprises a continuum from the true NOEL through the NOAEL for goitre development. The determined NOAEL is established principally by the studies of Knudsen et al. (2000) and the much larger study of Barrère et al. (2000) supported by Knudsen et al. (2002) and Cliff et al. (1986). The size of the cohorts used for the determination, particularly that of Barrère et al. (2987 subjects aged between 36 and 60 years), suggests a high confidence in the assessment and no need to apply any additional factor to account for the dose-response of the critical effect.

Quality of available human database (and animals) including:

a) In terms of Completeness, the combined database on the health effects of cyanides is one of the most complete known. In addition to the well established mechanism of action for both acute and chronic toxicity, there are three OECD guideline sub-chronic studies, that include more than one species, by different routes as well as extensive clinical experience and large epidemiology studies from more than one continent. Furthermore, the acute toxicity of cyanide, together with its steep dose response curve and proximity between onset of repeat-dose effects and acute lethality, makes the conduct of chronic studies in animals technically very challenging, if not impossible to conduct. Taking the weight-of-evidence approach, the database for developing the POD is sufficiently complete.

b) In terms of Consistency, the NOAELs from the 90d studies were broadly consistent; serum thiocyanate levels in the studies of Leuschner et al. (1991) and Monsanto (1985) in rats and Tewe and Maner (1980) in pigs were very similar to the human NOAEL for thiocyanate therapy and the level that did not lead to an increased incidence of goitre in a cassava-eating population that was not iodine deficient (Cliff et al., 1986) as well as electroplating workers without goitre (Banerjee, 1997). As there is a high level of consistency between the various animal and human studies, no additional factor to account for consistency is necessary.

c) In terms of Reliability, the animal studies are full OECD guideline studies of high reliability. The clinical data were sufficient to support use of thiocyanate as a therapeutic agent for over a decade and the epidemiology studies well conducted by multiple independent investigators using recognized approaches and published in peer reviewed journals. As such, the data are considered reliable and no additional factor to account for reliability is necessary.

d) In terms of Healthy worker effect, the only study relied on for the POD that could be subject to the healthy worker effect is that of Banerjee (1997) on electroplaters. As the other studies were conducted on individuals drawn from the public and many were outside the age range typically identified as representing the working age, this is not considered relevant. Hence, no additional factor to account for the healthy worker effect is necessary.

e) In terms of Study size, the human studies vary from several hundred to several thousand individuals. Collectively they represent a sufficiently large study size. Hence, no additional factor to account for study size is necessary.

In conclusion, there is an overall high confidence in the database and there is no need to apply additional uncertainty factors over and above an intraspecies uncertainty factor of 10 for the worker.

Derivation of DNELworker. – inhalation:

By applying a combined AF of 10* (10 to account for intraspecies differences in sensitivity (dietary iodide and normal renal) and 1 for interspecies uncertainty (as mainly derived from human data)) to the POD of 7.5 mg/m3 a DNELworker – inhalation of 0.75mgCN-/ m3, or 0.78 mgHCN/ m3, is therefore derived.

Derivation of DNELworker. – Dermal/Oral:

DNELworker – Dermal/Oral can be derived in a comparable way:

[CN-] mg/kg = ([SCN-] mg/l × 0.25 l/kgbw × 26) divided by

(0.8 × 3.9 × 58)

Where

0.25 l/kgbw = distribution volume of SCN

26 = molecular weight CN

0.8 = conversion of cyanide to thiocyanate: ~ 80%

3.9 = elimination of thiocyanate, half-life = 2.7 d, residence time (t0.5/ln 2) = 3.9 d

58 = molecular weight SCN-

The derived safe dermal/oral concentration of cyanide for humans exposed for 8 hours at the workplace of 0.54 mg/kg can be used as a departure point for the calculation of DNELworker.

By applying a combined AF of 10* (10 to account for intraspecies differences in sensitivity (dietary iodide and normal renal) and 1 for interspecies uncertainty (as mainly derived from human data)) to the POD of 7.5 mg/m3 a DNELworker – Dermal/Oral of 0.054 mgCN-/kg, or 0.054 mgHCN/kg, is therefore justified.

General Population - Hazard via inhalation route

Systemic effects

Long term exposure

Hazard assessment conclusion:

DNEL (Derived No Effect Level)

Value:

0.13 mg/m³

Most sensitive endpoint:

repeated dose toxicity

DNEL related information

Overall assessment factor (AF):

30

Acute/short term exposure

Hazard assessment conclusion:

DNEL (Derived No Effect Level)

Most sensitive endpoint:

acute toxicity

DNEL related information

Local effects

Acute/short term exposure

DNEL related information

General Population - Hazard via dermal route

Systemic effects

Long term exposure

Hazard assessment conclusion:

DNEL (Derived No Effect Level)

Value:

0.018 mg/kg bw/day

DNEL related information

Overall assessment factor (AF):

30

Acute/short term exposure

Hazard assessment conclusion:

DNEL (Derived No Effect Level)

Most sensitive endpoint:

acute toxicity

DNEL related information

General Population - Hazard via oral route

Systemic effects

Long term exposure

Hazard assessment conclusion:

DNEL (Derived No Effect Level)

Value:

0.018 mg/kg bw/day

Most sensitive endpoint:

repeated dose toxicity

DNEL related information

Overall assessment factor (AF):

30

Modified dose descriptor starting point:

NOAEL

Acute/short term exposure

Hazard assessment conclusion:

DNEL (Derived No Effect Level)

Most sensitive endpoint:

acute toxicity

DNEL related information

Additional information - General Population

In terms of General population the DNELworker can also be used as a basis for the derivation of the DNELgeneral population by applying a suitable AF.

ECETOC Technical Report 86 on Assessment factors suggest the use of an adjustment factor of 3 for extrapolation from DNELworker to DNELgeneral population whereas ECHA R8 Guidance (2008) suggests a factor of 2 (difference between factor 10 for intraspecies variation general population and 5 for the worker). Use of an AF of 3 is therefore sufficient to address any residual uncertainty regarding the completeness of the database and any residual concern relating to sensitive sub-populations within the general population such as pregnant women and their possible increased sensitivity to iodine insufficiency and its impact on thyroid hormone levels and developmental toxicity.

Derivation DNELgeneral population – inhalation:

ECETOC proposed that “A level of 15 μg SCN-/ml can be used as a starting point to derive a safe concentration of cyanide for humans, using the following equation.

[CN-] mg/m3 = ([SCN-] mg/l × 0.25 l/kgbw × 70 kgbw × 26) divided by

(10 m3 × 0.5 × 0.8 × 3.9 × 58)

Where

0.25 l/kgbw = distribution volume of SCN

70 kg = body weight

26 = molecular weight CN

20 m3 = respiratory volume during an 24hour day

0.5 = absorption of cyanide by the inhalation route: 50%

0.8 = conversion of cyanide to thiocyanate: ~ 80%

3.9 = elimination of thiocyanate, half-life = 2.7 d, residence time (t0.5/ln 2) = 3.9 d

58 = molecular weight SCN-

The derived safe inhalation concentration of cyanide for humans exposed for 24 hours of 3.75 mg/m3 can be used as a departure point for the calculation of DNELgeneral population. By applying a combined AF of 10* (10 to account for intraspecies differences in sensitivity (dietary iodide and normal renal) and 1 for interspecies uncertainty (as mainly derived from human data)) and an AF of 3 to address any residual uncertainty regarding the completeness of the database to the POD of 1.25 mg/m3 a DNELgeneral population – inhalation of 0.13mgCN-/ m3, or 0.13 mgHCN/ m3, is therefore derived.

Derivation of DNELgeneral population – Dermal/Oral:

By applying an AF of 3 to the DNELworker – Dermal/Oral of 0.054 mgHCN/kg a DNELgeneral population - Dermal/Oral of 0.018 mgHCN/kg is therefore derived.

The repeat dose toxicity of HCN was reviewed by ECETOC in the JACC report on Cyanides (No. 53, 2007) and concluded that an acceptable chronic level of exposure to CN- could be established using a weight-of-evidence approach that uses all the available repeat dose experimental studies in animals and human studies on the effects of exposure to cyanides/thiocyanate.